JPH09213008A - Decoding device and disk device having the same - Google Patents

Abstract

(57) Abstract: An error rate of output data is effectively reduced. A reproduction signal S _MO is PR by an equalization circuit 43.
(1,1) waveform equalization is performed, and PR (0.
5, 1, 0.5) Waveform equalization is performed. The identification / decoding circuits 45 to 47 identify / decode the output data of the equalization circuit 43 by different methods. Identification / decoding circuits 48 to 50 identify / decode the output data of the equalization circuit 44 by different methods. The output data (decoded data) of the identification / decoding circuits 45 to 50 is supplied to the calculation circuit 51, each error rate is calculated, and the calculation result is supplied to the syscon 16. The switching of the switch 52 is controlled by the system controller 16, and the identification / decoding circuit 4 is controlled by the switch 52.
The output data having the minimum error rate is taken out of the output data of 5 to 50, and RLL (1,7) demodulation processing and error correction processing are performed to obtain output data Dout.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a decoding device and a disk device having the same. More specifically, decoding that attempts to effectively reduce the error rate of the output data by decoding the input signal to be decoded by a plurality of different methods and selectively outputting the decoded data with the minimum error rate. The present invention relates to a device and a disk device having the device.

[0002]

2. Description of the Related Art In optical disc devices and the like, methods such as narrowing the track pitch and narrowing the code intervals before and after are adopted in order to increase the recording density. When the track pitch is narrowed, the crosstalk between tracks increases during reproduction, and the signal of the adjacent track leaks into the reproduced signal as noise. Further, when the code interval before and after is narrowed, the codes before and after are mixed at the time of reproduction, and inter-code interference occurs in the reproduced signal. Therefore, 0/1 of the reproduced signal is erroneously identified or decoded, and the error rate is deteriorated.

In optical fibers and information transmission media, the transmission band is determined by the material and structure thereof, but the upper limit of the transmission band is required due to the demand for higher speed communication.
Alternatively, when a signal in a band higher than that is transmitted, intersymbol interference similarly occurs, causing 0/1 identification and decoding of the received signal to be erroneous, resulting in deterioration of the error rate.

Therefore, in order to improve the error rate in the past, instead of simple binary identification and decoding by a threshold,
PRML (Partial Response Maximum Likelihood), partial response and viterbi
It has been proposed to use maximum likelihood decoding combined with decoding.

[0005]

The PRML described above has a plurality of methods, and the optimum method differs depending on the state of noise (including intersymbol interference) included in the reproduced signal. However, in a conventional device, for example, an optical disc device, one method of the greatest common divisor among them is adopted. Therefore, there is a problem that the error rate cannot be sufficiently improved.

Therefore, the present invention provides a decoding device capable of effectively reducing the error rate of output data and a disk device having the decoding device.

[0007]

SUMMARY OF THE INVENTION A decoding device according to the present invention provides a plurality of decoding means for decoding input signals to be decoded by different methods and an error rate of output data of the plurality of decoding means. An error rate calculating means for calculating each, and a data output means for selectively outputting the output data having the minimum error rate calculated by the error rate calculating means among the output data of the plurality of decoding means are provided. is there. For example, all or some of the plurality of decoding means are Viterbi decoding means. Further, for example, the input signal includes a signal corresponding to the specific pattern data for calculating the error rate.

Further, the disk device according to the present invention decodes the data reproduction means for reproducing the encoded data recorded on the disk-shaped recording medium and the reproduction signal outputted from the data reproduction means by different methods. A plurality of decoding means, an error rate calculating means for calculating an error rate of output data of the plurality of decoding means, respectively,
The data output means selectively outputs the output data having the minimum error rate calculated by the error rate calculation means from the output data of the plurality of decoding means.

An input signal, for example, a reproduction signal from a recording medium such as a magneto-optical disk on which encoded data is recorded, a received signal in digital transmission for transmitting encoded data, or the like is different depending on a plurality of decoding means. Will be decrypted with. In this case, the output data having the smallest error rate among the output data (decoded data) of the plurality of decoding means changes depending on the noise state of the input signal.

The error rate calculation means calculates the error rates of the output data of the plurality of decoding means, respectively. In this case, when the input signal includes a signal corresponding to the specific pattern data for calculating the error rate, for example, the error rate is calculated by counting the number of errors in the decoded data of the signal corresponding to the specific pattern data. Is calculated. Then, from the output data of the plurality of decoding means, the output data having the smallest error rate is selectively output from the data output means.

[0011]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows the configuration of a magneto-optical disk device 10 as an embodiment. This disk device 10 includes a spindle motor 12 for rotating a magneto-optical disk 11 at a constant angular velocity.
And a magnetic head 13 for generating an auxiliary magnetic field during recording.
And an optical head 14 including a laser diode, an objective lens, a photodetector, a preamplifier, and the like.
The magnetic head 13 and the optical head 14 are the magneto-optical disk 11
Are disposed so as to face each other.

The disk device 10 includes an optical head 1
4, a laser drive circuit 15 for driving the laser diode, and a CPU (central processing unit),
It has a system controller (hereinafter referred to as “syscon”) 16 for controlling the entire system. Here, the laser driving circuit 15 is supplied with the output laser power detection output S _DP from the optical head 14 and the power control signal from the syscon 16, and the optical head 14 is supplied with the power control signal.
Is controlled so that the power of the laser beam output from the laser diode becomes the optimum power at the time of recording and at the time of reproduction, respectively.

The recording data RD is supplied to the laser drive circuit 15 at the time of recording, as will be described later. Therefore, at the time of recording, the laser diode of the optical head 14 is driven by the laser drive circuit 15 so that the laser power changes according to the recording data RD. As a result, the recording data RD is magneto-optically recorded in the data section of the magneto-optical disk 11.

Further, the reproduction signal S _MO from the data section (MO area) of the magneto-optical disk 11 is obtained from the optical head 14 at the time of reproduction, and the pre-format section (ROM of the magneto-optical disk 11 at the time of recording and reproduction). The reproduction signal S _RF from the area) is obtained. Further, the optical head 14
From the tracking error signal obtained by the conventionally known detection method E _T, and a focus error signal E _F is output.

FIG. 2 shows the recording format of each sector of the magneto-optical disk. At the beginning of each sector, a pre-formatted section in which addresses and the like are recorded in advance by pits is arranged, and following this pre-formatted section, an ALPC (Auto Laser Power) as a test section for controlling the output laser power level. Control) section is arranged. Further, a data section is arranged following the ALPC section. Then, in the present embodiment, the specific pattern data ERR for calculating the error rate is magneto-optically recorded in the first fixed area of the data portion.

The disk device 10 also has a servo circuit 17 having a CPU. In the servo circuit 17,
Error signal E _T outputted from the optical head 14, E _F is supplied. The operation of the servo circuit 17 is controlled by the syscon 16. The servo circuit 17 controls an actuator 18 including a tracking coil, a focus coil, and a linear motor for moving in the radial direction to perform tracking and focus servo of the optical head 14 and to move the optical head 14 in the radial direction. Movement is controlled.

The disk device 10 includes an optical head 1
4 reproduced signal S from the preformat section
PLL (Phase- for reproducing clock CK _RF from _RF
Locked Loop) circuit 21 and the reproduction signal S _RF to the clock C
A / D for converting to digital signal using K _RF
It has an (Analog-to-digital) converter 22 and an address decoder 23 for obtaining address data AD from an output signal of the A / D converter 22. The address data AD obtained by the address decoder 23 is syscon 1
6 and is used for access control during recording and reproduction.

Further, the disk device 10 is provided with a data interface 31 for receiving input data Din from the outside, for example, a host computer, and for adding an error correction code to the input data Din received by the data interface 31. It has an ECC encoder 32 and a channel encoder 33 that performs RLL (1,7) modulation processing as digital modulation processing on the output data of the ECC encoder 32. Where ECC
The sector data recorded in the data portion of each sector is sequentially output from the encoder 32, and the channel encoder 33
The data corresponding to the specific pattern data ERR recorded in the first fixed area of the data portion for calculating the error rate as described above is inserted at the beginning of each sector data.

Further, the disk device 10 has an intermediate sequence, that is, NRZI (Non-Return-t) in order to avoid propagation of a code error at the time of identifying the output data of the channel encoder 33.
An o-Zero-Inverted) code is included in the pre-coding circuit 34. The recording data RD output from the precoding circuit 34 is supplied to the laser driving circuit 15 and recorded in the data section of the magneto-optical disk 11 as described above.

The disk device 10 includes an optical head 1
4 the reproduced signal S from the data section output from_MOMore black
CK_MOPLL circuit 41 for reproducing the
No. S _MOClock CK_MOTo convert to a digital signal
A / D converter 42 for converting and this A / D converter 4
Playback signal S converted to digital signal in 2_MOAgainst
Waveform equalization circuit 4 for performing PR (1, 1) waveform equalization
3 and its reproduction signal S _MOFor PR (0.5, 1,
0.5) and a waveform equalization circuit 44 for performing waveform equalization
Have.

By the way, the reproduction signal S _MO is band-limited by an MTF (Modulation Transfer Function) determined by the optical system, and the signal amplitude becomes smaller as the frequency becomes higher. FIG. 3 shows an example of the MTF characteristic. This M
Due to the TF characteristic, a signal having a higher frequency, that is, a signal having a smaller pulse width (sign inversion width) has a smaller amplitude. In such a signal with a small amplitude, it is difficult to distinguish 0/1. Therefore, a process for approximating the original recorded waveform by passing through a high-frequency emphasis filter having a characteristic opposite to MTF (waveform equalization) )
Is performed. There are several methods for waveform equalization, and there is a combination with the identification processing described below. However, in the magneto-optical disk device, the PR (1,1) and PR (0.5,1,0) described above are generally used because the frequency characteristic is close to the transfer characteristic of the magneto-optical disk device and the circuit is simplified. .5) are used.

The waveform equalization circuits 43 and 44 are realized by a transversal filter as shown in FIG. This transversal filter has, for example, a configuration in which a plurality of delay lines DL (delay time T ₀ ) are connected in series, and the output of each tap is multiplied by a tap coefficient to obtain the sum. Here, the number of taps is 2n + 1, the output of each tap is pb [-n] to pb [n], and the tap coefficient is C [-n] to.
When C [n] is set, the output data Y [k] is as shown in equation (1). In this case, the filter characteristics can be changed by changing the values of the tap coefficients C [-n] to C [n].

[0023]

[Equation 1]

Returning to FIG. 1, the disk device 10
Are identification / decoding circuits 45 to 47 for performing 0/1 identification and decoding for the output data of the waveform equalization circuit 43, and 0/1 identification for the output data of the waveform equalization circuit 44. It has identification / decoding circuits 48 to 50 for performing decoding. These identification / decoding circuits 45 to 47 and 48 to 50 are
Each uses Viterbi decoding as maximum likelihood decoding. For Viterbi decoding, in addition to ternary 2-state Viterbi decoding (VD32) in which the concept of state transition is introduced into simple ternary processing, ternary 4-state considering the law of connection of the preceding and following codes There are several methods such as Viterbi decoding (VD34) and 4-value 4-state Viterbi decoding (VD44). For example, the identification / decoding circuits 45 and 48 use ternary 2-state Viterbi decoding, the identification / decoding circuits 46 and 49 use ternary 4-state Viterbi decoding, and the identification / decoding circuit 47,
In 50, 4-value 4-state Viterbi decoding is used.

The three-value two-state Viterbi decoding will be briefly described. That is, in the combination of NRZI code and PR (1,1) waveform equalization, the reproduced signal after PR (1,1) equalization is c [k], and the decoding result from it is called “a [k]”. For example, the state transition having two states S0 and S1 is performed as shown in FIG. Therefore, as shown in FIG.
For [k], set two thresholds LH and LL, and
The area of A, 0, -A is determined, and the result of repeating the most likely transition between S0 and S1 is decoded based on the value at the identification time point t1.

Next, the three-value four-state Viterbi decoding will be briefly described. That is, in the data after the RLL (1,7) modulation processing, 1 does not occur twice consecutively. That is, in the state transition diagram of the three-value two-state shown in FIG.
After S0 → S1, it always becomes S1, and after S1 → S0 always becomes S0. Therefore, as shown in FIG. 7, a state transition having four states S0 to S3 is performed. Therefore, as in the case of the three-value two-state Viterbi decoding, the thresholds LH and LL are set as shown in FIG. 6, and the most likely transition between S0 to S3 is made based on the value at the identification time t1. Decode the repeated result.

Next, the 4-value 4-state Viterbi decoding will be briefly described. That is, NRZI code and PR (0.5,
1, 0.5) PR in combination of waveform equalization
The reproduced signal (eye pattern) after (0.5, 1, 0.5) equalization is as shown in FIG. When only 1 or 0 is continuous, the amplitude is ± A, and 1 or 0 is 3
When three or more are continuous, the amplitude is ± (A + B), and 4-value discrimination is possible.

NRZI code and PR (0.5, 1, 0.
5) Considering the condition that 1 does not occur twice consecutively in the data after combining the waveform equalization and RLL (1,7) modulation processing, PR (0.5,1,
0.5) If the reproduced signal after equalization is c [k] and the decoding result from it is "a [k]", there is a state transition having four states S0 to S3 as shown in FIG. . Therefore, as shown in FIG. 8, three threshold values L are set for the reproduction signal c [k].
H, 0, LL are set and + (A + B), + A, -A,-
The area of (A + B) is determined, and the result of repeating the most likely transition between S0 to S3 is decoded based on the value at the identification time t2.

Further, returning to FIG. 1, the disk device 10
Has an error rate calculation circuit 51 for calculating the error rate of the output data (decoded data) of the identification / decoding circuits 45 to 50, respectively. Here, in the error rate calculation circuit 51, the identification / decoding circuit 45 corresponding to the reproduction signal S _MO of the specific pattern data ERR recorded in the first fixed area of the data portion of the magneto-optical disk 11 as described above.
The number of errors in output data (correct data is known) of .about.50 is counted, and the error rate is calculated for each sector. Then, the calculation result of the error rate calculated by the error rate calculation circuit 51 is supplied to the syscon 16.

The disk device 10 also has a changeover switch 52 for selectively taking out any of the output data of the identification / decoding circuits 45-50. In this case, the output terminals of the identification / decoding circuits 45 to 50 are connected to the fixed terminals on the a to f sides of the changeover switch 52, respectively. The switching of the changeover switch 52 is controlled by the syscon 16, and is controlled for each sector so that the output data having the minimum error rate among the output data of the identification / decoding circuits 45 to 50 is output from the movable terminal.

Further, the disk device 10 includes a channel decoder 53 for demodulating the output data of the changeover switch 52, and an ECC encoder 54 for performing error correction processing on the output data of the channel decoder 53.
The ECC encoder 54 has a data interface 55 for transmitting the error-corrected data output from the ECC encoder 54 to the outside, for example, a host computer.

The operation of the magneto-optical disk device 10 shown in FIG. 1 will be described. First, the operation during recording will be described. At the time of recording, after input data Din supplied from the outside, for example, from a host computer, is received by the data interface 31, an error correction code is added by the ECC encoder 32. Then, sector data recorded in the data portion of each sector is sequentially output from the ECC encoder 32 and supplied to the channel encoder 33. in this case,
At the beginning of each sector data, data corresponding to the specific pattern data ERR recorded in the first fixed area of the data portion is inserted and supplied in order to calculate the error rate.

Further, the channel encoder 33 performs RLL (1,7) modulation processing as digital modulation processing. Then, the modulation data output from the channel encoder 33 is converted into an NRZI code by the precode circuit 34 and supplied to the laser drive circuit 15 as recording data RD. As a result, the recording data RD is magneto-optically recorded on the data portion of the magneto-optical disk 11. In this case, the specific pattern data ERR for calculating the error rate is recorded in the first fixed area of the data part.

Next, the operation during reproduction will be described. Again
At the time of production, the magneto-optical disk 11 is
From the data portion of the reproduction signal S_MOIs played. This playback message
No. S _MOIs supplied to the waveform equalization circuit 43 and PR (1,1)
Is supplied to the waveform equalization circuit 44 as the waveform is equalized.
Then, PR (0.5, 1, 0.5) waveform equalization is performed.
You. Then, for the output data of the waveform equalization circuit 43,
The identification / decoding circuit 45 performs three-value two-state Viterbi decoding,
The identification / decoding circuit 46 performs 3-value 4-state Viterbi decoding,
Further, 4-value 4-state Viterbi decoding is performed by the identification / decoding circuit 47.
Will be Similarly, for the output data of the waveform equalization circuit 44,
Then, the identification / decoding circuit 48 performs 3-valued 2-state Viterbi decoding.
Then, the identification / decoding circuit 49 performs 3-value 4-state Viterbi decoding.
Further, 4-value 4-state Viterbi decoding is performed by the identification / decoding circuit 50.
Is performed.

The output data (decoded data) of the identification / decoding circuits 45 to 50 is supplied to the error rate calculation circuit 51, each error rate is calculated for each sector, and the calculation result is supplied to the syscon 16. To be done. Then, the changeover switch 52 is controlled for each sector under the control of the system controller 16, and the output data having the minimum error rate is taken out from the output data of the identification / decoding circuits 45 to 50 from the changeover switch 52.

Then, the data taken out by the changeover switch 52 is demodulated by the channel decoder 53 and then E
The error correction processing is performed by the CC encoder 54, and the data subjected to the error correction processing is sent from the data interface 55 to the outside, for example, the host computer as output data Dout.

As described above, in the present embodiment, a combination of a plurality of waveform equalization circuits 43, 44 and the identification / decoding circuits 45 to 50 is provided in the reproduction system, and the output data of the identification / decoding circuits 45 to 50 is provided. Of these, the output data having the smallest error rate is taken out from the changeover switch 52 and adopted, and the error rate of the output data can be effectively reduced. Further, even if the reproduced signal S _MO has more noise than the conventional one, the error rate can be suppressed to the same as the conventional one. Therefore, if the same error rate as the conventional one is obtained,
It is possible to further improve the recording density of the magneto-optical disk 11.

In the error rate calculation circuit 51, the output data of the identification / decoding circuits 45 to 50 corresponding to the reproduction signal S _MO of the specific pattern data ERR recorded in the first fixed area of the data portion of the magneto-optical disk 11 is output. The number of errors is counted to calculate the error rate, and the error rate can be calculated easily.

Even if not mentioned above, the optimum waveform equalization circuit and identification /
The decoding circuit differs depending on the S / N, the frequency component of noise, the recording position such as the inner and outer circumferences of the disc, and the dispersion of individual devices. As described above, controlling the switching of the changeover switch 52 according to the error rate of the output data of the identification / decoding circuits 45 to 50 means that the above-described S / N, noise frequency components, inner and outer circumferences of the disk, etc. are recorded. position,
This means that it is considered to be more suitable than the one in which the switching control is performed based on, for example, the S / N and the frequency component of noise, etc.

The error rate calculation circuit 51 calculates the error rate of the output data of the identification / decoding circuits 45 to 50 for each sector, and the switching of the changeover switch 52 is controlled by the syscon 16 for each sector. N,
Immediate response to changes in noise frequency components, recording position, etc.

In the above embodiment, the combination of the two types of waveform equalization circuits 43 and 44 and the three types of identification / decoding circuits 45 to 50 is shown, but the present invention is not limited to this. Of course. Further, in the above-described embodiment, the one applied to the magneto-optical disk device 10 is shown, but the present invention can be similarly applied to other disk devices, and further to a digital signal transmission device.

[0042]

According to the present invention, the input signal to be decoded is decoded by a plurality of different methods, and the decoded data having the minimum error rate is selectively output. It can be effectively reduced. Further, even if the input signal has more noise than the conventional one, the error rate can be suppressed to the same as the conventional one. Therefore, if the same error rate as the conventional one is obtained, for example, the recording density on the recording medium can be improved. It will be possible.

When the input signal includes a signal corresponding to the specific pattern data for calculating the error rate, the error rate is calculated by counting the number of errors in the decoded data of the signal corresponding to the specific pattern data. By calculating, the error rate can be easily calculated.

[Brief description of drawings]

FIG. 1 is a block diagram showing a magneto-optical disk device as an embodiment.

FIG. 2 is a diagram showing a recording format of each sector.

FIG. 3 is a diagram showing MTF characteristics.

FIG. 4 is a diagram showing a configuration of a transversal filter that constitutes a waveform equalization circuit.

FIG. 5 is a diagram showing a state transition diagram in three-value two-state Viterbi decoding (VD32).

FIG. 6 is a diagram showing an eye pattern of a reproduced signal when NRZI code and PR (1,1) waveform equalization are combined.

FIG. 7 is a diagram showing a state transition diagram in three-value four-state Viterbi decoding (VD34).

FIG. 8 is a diagram showing an eye pattern of a reproduced signal when NRZI code and PR (0.5, 1, 0.5) waveform equalization are combined.

FIG. 9 is a diagram showing a state transition diagram in 4-value 4-state Viterbi decoding (VD44).

[Explanation of symbols]

10 ... Magneto-optical disk device, 11 ... Magneto-optical disk, 14 ... Optical head, 15 ... Laser drive circuit, 16 ... System controller, 32 ... EC
C encoder, 33 ... Channel encoder, 34 ...
..Precode circuits, 42 ... A / D converters, 43,
44 ... Waveform equalization circuit, 45-50 ... Identification / decoding circuit, 51 ... Error rate calculation circuit, 52 ... Changeover switch, 53 ... Channel decoder, 54 ...
ECC decoder

Claims (8)

[Claims] 1. A plurality of decoding means for decoding input signals to be decoded by different methods, an error rate calculating means for calculating an error rate of output data of the plurality of decoding means, and the plurality of decoding means. A decoding device, comprising: data output means for selectively outputting output data having a minimum error rate calculated by the error rate calculation means, out of output data of the conversion means. 2. The decoding apparatus according to claim 1, wherein all or some of the plurality of decoding means are Viterbi decoding means. 3. The decoding device according to claim 1, wherein the input signal includes a signal corresponding to specific pattern data for calculating the error rate. 4. The decoding device according to claim 1, wherein the input signal is a reproduction signal from a recording medium on which encoded data is recorded. 5. The decoding device according to claim 4, wherein specific pattern data for calculating the error rate is recorded in a predetermined area of the recording medium. 6. The decoding device according to claim 1, wherein the input signal is a reception signal in digital transmission for transmitting encoded data. 7. Data reproducing means for reproducing encoded data recorded on a disk-shaped recording medium, and a plurality of decoding means for decoding reproduced signals output from the data reproducing means by different methods, respectively. Error rate calculation means for respectively calculating error rates of output data of a plurality of decoding means, and output data having a minimum error rate calculated by the error rate calculation means among the output data of the plurality of decoding means And a data output unit that selectively outputs the disk device. 8. The disk device according to claim 7, wherein the disk-shaped recording medium is a magneto-optical disk.